Formatting changes (5f9a76d5) · Commits · Centre for Testing and Interoperability / Markdown specifications development / Specification tools

grid_table_filter.rb

+487 −480

Original line number	Diff line number	Diff line
		@@ -34,6 +34,12 @@ module Banzai

		require 'logger'

		# Add these regex constants at the top of the file, after the require statement
		GRID_TABLE_SEPARATOR = /^\s\+([-:=]+\+)+\s$/
		GRID_TABLE_HEADER_SEPARATOR = /^\s\+([=:]+\+)+\s$/
		GRID_TABLE_BODY_SEPARATOR = /[^\n]\+([:-]+\+)+[^\n]$/
		GRID_TABLE_BODY_SEPARATOR_LINE = /^[-:]+$/

		class Cell
		attr_accessor :content, :rowspan, :colspan, :colspan_adjusted, :alignment, :position, :list_flag

		@@ -49,20 +55,18 @@ module Banzai

		def set_alignment(default_alignments, header_delimiter_positions)
		header_delimiter_index = 0
		while header_delimiter_index < default_alignments.length && @position > header_delimiter_positions[header_delimiter_index]
		while header_delimiter_index < default_alignments.length &&
		@position > header_delimiter_positions[header_delimiter_index]
		header_delimiter_index += 1
		end

		if header_delimiter_index < default_alignments.length
		raise "Invalid table formatting" unless header_delimiter_index < default_alignments.length
		if @position < header_delimiter_positions[header_delimiter_index]
		@alignment = default_alignments[header_delimiter_index]
		elsif @position == header_delimiter_positions[header_delimiter_index]
		@alignment = default_alignments[header_delimiter_index]
		header_delimiter_index += 1
		end
		else
		raise "Invalid table formatting"
		end
		end
		end # end of class Cell

		@@ -102,21 +106,10 @@ module Banzai
		end
		end # end of class RowTracker

		# Add these regex constants at the top of the file, after the require statement
		GRID_TABLE_SEPARATOR = /^\s\+([-:=]+\+)+\s$/
		GRID_TABLE_HEADER_SEPARATOR = /^\s\+([=:]+\+)+\s$/
		GRID_TABLE_BODY_SEPARATOR = /[^\n]\+([:-]+\+)+[^\n]$/
		GRID_TABLE_BODY_SEPARATOR_LINE = /^[-:]+$/

		def parse_pandoc_table_with_spans(pandoc_table)
		# Split the input into lines
		lines = pandoc_table.strip.split("\n").map(&:strip)

		# Helper method to detect separator lines
		def is_separator(line)
		def separator?(line)
		GRID_TABLE_SEPARATOR.match?(line)
		end

		# Helper method to handle content in cells
		def handling_content(cell, content)
		if cell.content.nil?
		@@ -133,11 +126,9 @@ module Banzai
		else
		cell.content = content.strip.gsub(/\\\s*$/, "\n")
		end
		else
		if content.strip.start_with?("- ")
		unless cell.list_flag
		cell.content += "\n"
		end
		elsif content.strip.start_with?("- ")
		cell.content += "\n" unless cell.list_flag

		cell.list_flag = true
		cell.content += "#{content.strip}\n"
		elsif cell.list_flag && !content.strip.empty?
		@@ -150,7 +141,7 @@ module Banzai
		content = content.strip.gsub(/\\\s*$/, "\n")
		cell.content += " #{content}"
		end
		end

		cell
		end

		@@ -159,12 +150,14 @@ module Banzai
		(column_index...number_of_parts).each do \|j\|
		delimiter_start = nil
		col_i = column_index

		until delimiter_start
		delimiter_start = col_i > 0 ? row[col_i - 1].position : 0
		col_i -= 1
		end

		positions = ["\|", "+"].map do \|delimiter\|
		delimiters = ['\|', '+']
		positions = delimiters.filter_map do \|delimiter\|
		pos = line[delimiter_start + 1..-1]&.index(delimiter)
		pos ? pos + delimiter_start + 1 : nil
		end.compact
		@@ -173,6 +166,7 @@ module Banzai

		if position && position > delimiter_positions[j]
		row[column_index].colspan += 1

		if position == delimiter_positions[-1]
		colspan_allocated = row[column_index].colspan
		row[column_index].colspan += number_of_columns - colspan_allocated - column_index
		@@ -183,11 +177,20 @@ module Banzai
		break
		end
		end

		row[column_index]
		end

		# rubocop:disable Metrics/AbcSize -- PoC
		# rubocop:disable Metrics/CyclomaticComplexity -- PoC
		# rubocop:disable Metrics/PerceivedComplexity -- PoC
		def parse_pandoc_table_with_spans(pandoc_table)

		# Split the input into lines
		lines = pandoc_table.strip.split("\n").map(&:strip)

		# Retrieve separator indices
		separator_indices = lines.each_index.select { \|i\| is_separator(lines[i]) }
		separator_indices = lines.each_index.select { \|i\| separator?(lines[i]) }

		raise "No valid separators found in the provided Pandoc table." if separator_indices.empty?

		@@ -198,7 +201,7 @@ module Banzai
		# Determine delimiter positions
		separator_index_max_columns = separator_indices.find { \|i\| lines[i].count("+") - 1 == number_of_columns }
		number_of_columns.times do \|j\|
		start_pos = j.zero? ? 0 : delimiter_positions[j - 1]
		start_pos = j == 0 ? 0 : delimiter_positions[j - 1]
		pos = lines[separator_index_max_columns][start_pos + 1..-1]&.index("+")
		delimiter_positions << (pos ? pos + start_pos + 1 : -1)
		end
		@@ -212,7 +215,8 @@ module Banzai

		# Determine header delimiter positions
		separator_indices.each do \|index\|
		if GRID_TABLE_HEADER_SEPARATOR.match?(lines[index])
		next unless GRID_TABLE_HEADER_SEPARATOR.match?(lines[index])

		has_header = true
		header_separator_index = index
		parts = lines[index].strip.delete_prefix("+").split("+")
		@@ -226,28 +230,29 @@ module Banzai
		'align="center"'
		end

		start_pos = part_index.zero? ? 0 : header_delimiter_positions[part_index - 1]
		start_pos = part_index == 0 ? 0 : header_delimiter_positions[part_index - 1]
		pos = lines[index][start_pos + 1..-1]&.index("+")
		header_delimiter_positions << (pos ? pos + start_pos + 1 : -1)
		end
		break
		end
		end

		# Process table body (including rows belonging to header as they are processed in the same way)
		data_rows = []

		(separator_indices.length - 1).times do \|row\|
		rows = []
		rows_tracker = nil
		in_data_row = false
		start, end_idx = separator_indices[row], separator_indices[row + 1] # Lines between separators including separator line start as it gives information about the number of columns of the row
		start = separator_indices[row]
		end_idx = separator_indices[row + 1]
		row_lines = lines[start...end_idx]

		next if row_lines.empty?

		row_lines.each do \|line\|
		# First line (normally a separator) of each block
		if is_separator(line) && !in_data_row
		if separator?(line) && !in_data_row
		in_data_row = true
		parts = line.strip.delete_prefix("+").split("+")
		delimiter_index = 0
		@@ -262,17 +267,17 @@ module Banzai
		rows[-1][i].position = delimiter_index
		rows[-1][i].set_alignment(default_alignments, header_delimiter_positions)

		while delimiter_index > delimiter_positions[i]
		i += 1
		end
		i += 1 while delimiter_index > delimiter_positions[i]
		i += 1
		end
		# Lines in a block
		elsif in_data_row
		# Regular data row or partial separator
		if GRID_TABLE_BODY_SEPARATOR.match?(line) # Partial separator
		cells_content = line.strip.delete_prefix("\|").delete_prefix("+")
		.delete_suffix("\|").delete_suffix("+").split(/[\\|\+]/)
		cells_content = line.strip
		.delete_prefix("\|")
		.delete_prefix("+")
		.split(/[\\|\+]/)

		rows << Row.new(number_of_columns)
		aux_delimiter_index = 0
		@@ -285,16 +290,16 @@ module Banzai
		rows[-1][auxiliar_cell_index].position = aux_delimiter_index
		rows[-1][auxiliar_cell_index].set_alignment(default_alignments, header_delimiter_positions)

		while aux_delimiter_index > delimiter_positions[auxiliar_cell_index]
		auxiliar_cell_index += 1
		end
		auxiliar_cell_index += 1 while aux_delimiter_index > delimiter_positions[auxiliar_cell_index]

		auxiliar_cell_index += 1
		end

		if cells_content.length <= number_of_columns
		raise "More cells than columns found" unless cells_content.length <= number_of_columns

		column_index = 0
		maxRowTracker = rows_tracker.maxValue
		cells_content.each_with_index do \|content, i\|
		cells_content.each_with_index do \|content, _i\|
		if GRID_TABLE_BODY_SEPARATOR_LINE.match?(content) # Separator - split row
		rows_tracker[column_index] = maxRowTracker + 1
		rows[rows_tracker[column_index]][column_index].list_flag = false
		@@ -306,6 +311,7 @@ module Banzai
		#rows_tracker[column_index + column_forward - 1] += 1 if column_forward > 1
		end
		end

		column_index += column_forward
		else # Regular cell in Partial separator line
		rows[rows_tracker[column_index]][column_index] =
		@@ -315,33 +321,39 @@ module Banzai
		unless rows[rows_tracker[column_index]][column_index].colspan_adjusted
		rows[rows_tracker[column_index]][column_index].colspan_adjusted = true
		rows[rows_tracker[column_index]][column_index] =
		adjust_colspan(rows[rows_tracker[column_index]], column_index, number_of_columns,
		line, number_of_columns, delimiter_positions)
		adjust_colspan(rows[rows_tracker[column_index]],
		column_index,
		number_of_columns,
		line,
		number_of_columns,
		delimiter_positions)
		end

		if rows[rows_tracker[column_index]][column_index].position >= delimiter_positions[column_index]
		colspan = rows[rows_tracker[column_index]][column_index].colspan
		column_index += colspan.zero? ? 1 : colspan
		column_index += (colspan == 0 ? 1 : colspan) # rubocop:disable Metrics/BlockNesting -- PoC
		end
		end
		end
		else
		raise "More cells than columns found"
		end

		else # Data row
		cells_content = line.strip.delete_prefix("\|").delete_suffix("\|").split(/\\|/)
		cells_content = line.strip.delete_prefix("\|").split(/\\|/)
		column_index = 0

		if cells_content.length < number_of_columns
		cells_content.each_with_index do \|content, i\|
		cells_content.each_with_index do \|content, _i\|
		rows[rows_tracker[column_index]][column_index] =
		handling_content(rows[rows_tracker[column_index]][column_index], content)

		unless rows[rows_tracker[column_index]][column_index].colspan_adjusted
		rows[rows_tracker[column_index]][column_index].colspan_adjusted = true
		rows[rows_tracker[column_index]][column_index] =
		adjust_colspan(rows[rows_tracker[column_index]], column_index, number_of_columns,
		line, number_of_columns, delimiter_positions)
		adjust_colspan(rows[rows_tracker[column_index]],
		column_index,
		number_of_columns,
		line,
		number_of_columns,
		delimiter_positions)
		end

		if rows[rows_tracker[column_index]][column_index].position >= delimiter_positions[column_index]
		@@ -376,35 +388,26 @@ module Banzai
		row.each do \|cell\|
		next if cell.content.nil?

		["**", "__"].each do \|bold_chars\|
		delimters = ['**', '__']
		delimters.each do \|bold_chars\|
		while cell.content.include?(bold_chars)
		cell.content = cell.content.sub(bold_chars, "<strong>")
		.sub(bold_chars, "</strong>")
		end
		end

		while cell.content.include?("_") && !cell.content.include?("\\_")
		while cell.content.include?("_") && cell.content.exclude?("\\_")
		cell.content = cell.content.rstrip.sub("_", "<i>").sub("_", "</i>")
		end

		while cell.content.include?("\\_")
		cell.content = cell.content.rstrip.sub("\\_", "_")
		end
		end
		end
		end
		cell.content = cell.content.rstrip.sub("\\_", "_") while cell.content.include?("\\_")

		# Convert newlines to HTML breaks
		[header_rows, data_rows].each do \|rows\|
		rows.each do \|row\|
		row.each do \|cell\|
		cell.content = cell.content&.gsub("\n", "<br />")
		end
		end
		end

		# Validate grid correctness
		[header_rows, data_rows].each do \|rows\|
		forward_rowspan = []

		rows.each_with_index do \|row, row_index\|
		@@ -413,16 +416,14 @@ module Banzai

		row.each_with_index do \|cell, cell_index\|
		sum += cell.colspan
		if row_index > 0 && cell.colspan.zero?
		if forward_rowspan[cell_index].positive?
		sum += 1
		end

		if row_index > 0 && cell.colspan == 0
		sum += 1 if forward_rowspan[cell_index] > 0

		forward_rowspan[cell_index] -= 1
		end

		if forward_rowspan[cell_index].zero? && cell.rowspan > 1
		forward_rowspan[cell_index] = cell.rowspan - 1
		end
		forward_rowspan[cell_index] = cell.rowspan - 1 if forward_rowspan[cell_index] == 0 && cell.rowspan > 1
		end

		raise "Grid table not converted properly" unless sum == number_of_columns
		@@ -436,9 +437,10 @@ module Banzai
		begin
		grid_header, grid_body = parse_pandoc_table_with_spans(pandoc_table)
		rescue StandardError => e
		logger = Logger.new(STDOUT)
		logger = Logger.new($stdout)
		logger.error("Grid table could not be generated: #{e.message}")
		return "HTML TABLE COULD NOT BE GENERATED FROM MARKDOWN GRID TABLE. CHECK LOGS FILE"

		"HTML TABLE COULD NOT BE GENERATED FROM MARKDOWN GRID TABLE. CHECK LOGS"
		else
		html = "<table>\n"
		has_header = false
		@@ -460,20 +462,21 @@ module Banzai
		next if cell.rowspan == 0 \|\| cell.colspan == 0

		# Prepare content, in case there's a list
		if matches = cell.content&.scan(/\s([-+]\|\d+\.)\s+([^<]+?)(?=<br \/>\|$)/)
		matches = cell.content&.scan(%r{\s([-+]\|\d+\.)\s+([^<]+?)(?=<br />\|$)})
		if matches
		list = "<ul>"
		matches.each do \|match\|
		list += "<li>#{match[1]}</li>"
		end
		list += "</ul>"
		cell.content = cell.content.gsub(/(\s([-+]\|\d+\.)\s+[^<]+?<br \/>)+/, list)
		cell.content = cell.content.gsub(%r{(\s([-+]\|\d+\.)\s+[^<]+?<br />)+}, list)
		# Enforce left alignment if cell contains a list
		cell.alignment = 'align="left"'
		end

		rowspan = cell.rowspan > 1 ? %( rowspan="#{cell.rowspan}") : ""
		colspan = cell.colspan > 1 ? %( colspan="#{cell.colspan}") : ""
		html += %{ <th#{rowspan}#{colspan} #{cell.alignment}>#{cell.content}</th>\n}
		html += %( <th#{rowspan}#{colspan} #{cell.alignment}>#{cell.content}</th>\n)
		end
		html += " </tr>\n"
		end
		@@ -486,20 +489,21 @@ module Banzai
		row.each do \|cell\|
		next if cell.rowspan == 0 \|\| cell.colspan == 0

		if matches = cell.content&.scan(/\s([-+]\|\d+\.)\s+([^<]+?)(?=<br \/>\|$)/)
		matches = cell.content&.scan(%r{\s([-+]\|\d+\.)\s+([^<]+?)(?=<br />\|$)})
		if matches
		list = "<ul>"
		matches.each do \|match\|
		list += "<li>#{match[1]}</li>"
		end
		list += "</ul>"
		cell.content = cell.content.gsub(/(\s([-+]\|\d+\.)\s+[^<]+?<br \/>)+/, list)
		cell.content = cell.content.gsub(%r{(\s([-+]\|\d+\.)\s+[^<]+?<br />)+}, list)
		# Enforce left alignment if cell contains a list
		cell.alignment = 'align="left"'
		end

		rowspan = cell.rowspan > 1 ? %( rowspan="#{cell.rowspan}") : ""
		colspan = cell.colspan > 1 ? %( colspan="#{cell.colspan}") : ""
		html += %{ <td#{rowspan}#{colspan} #{cell.alignment}>#{cell.content}</td>\n}
		html += %( <td#{rowspan}#{colspan} #{cell.alignment}>#{cell.content}</td>\n)
		end
		html += " </tr>\n"
		end
		@@ -509,6 +513,9 @@ module Banzai
		html
		end
		end # end of def generate_html_table_with_spans
		# rubocop:enable Metrics/PerceivedComplexity
		# rubocop:enable Metrics/CyclomaticComplexity
		# rubocop:enable Metrics/AbcSize

		def call
		return @text unless MarkdownFilter.glfm_markdown?(context)